Temozolomide in combination with BCNU before and after radiotherapy in patients with inoperable newly diagnosed glioblastoma multiforme

Original article Annals of Oncology 16: 1177 1184, 2005 doi:10.1093/annonc/mdi225 Published online 27 April 2005 Temozolomide in combination with BCNU before and after radiotherapy in patients with inoperable newly diagnosed glioblastoma multiforme M. Barrié 1, C. Couprie 2, H. Dufour 1, D. Figarella-Branger 3, X. Muracciole 4, K. Hoang-Xuan 5, D. Braguer 2, P. M. Martin 6, J. C. Peragut 1, F. Grisoli 1 & O. Chinot 1,6 * 1 Unité de Neuro-Oncologie, Service de Neurochirurgie, CHU Timone, Assistance Publique Hôpitaux de Marseille, 2 Service de Pharmacie, CHU Timone, 3 Laboratoire d Anatomie Pathologie, CHU Timone, 4 Service de Radiothérapie, CHU Timone, 5 Département de Neurologie, CHU Pitié Salpêtrière, Paris; 6 Laboratoire de Cancérologie Expérimentale, INSERM EMI 0359, Université de la Méditerranée, Faculté de Médecine de Marseille, Marseille, France Received 11 August 2004; revised 25 February 2005; accepted 28 February 2005 Background: The aim of this study was to evaluate the efficacy and safety of carmustine (BCNU) in combination with temozolomide as first-line chemotherapy before and after radiotherapy (RT) in patients with inoperable, newly diagnosed glioblastoma multiforme (GBM). Patients and methods: Forty patients were treated with BCNU (150 mg/m 2 ) on day 1 and temozolomide (110 mg/m 2 /day) on days 1 through 5 of each 42-day cycle for up to four cycles prior to conventional RT (2 Gy fractions to a total of 60 Gy). After RT, BCNU + temozolomide was administered for four additional cycles or until progression. The primary end point was response rate; secondary end points included progression-free survival (PFS); overall survival (OS) and safety. Results: Sixty per cent of patients completed four cycles of neo-adjuvant BCNU + temozolomide. Objective response rate (intention-to-treat) was 42.5% (95% confidence interval 27% to 58%), including two (5%) complete and 15 (37.5%) partial responses. In the eligible population (n = 37) the objective response rate was 46%. Nine (24%) patients had stable disease and 14 (35%) had progressive disease. Median PFS and OS were 7.4 and 12.7 months, respectively. Age was the only significant prognostic factor and tumor location (lobar versus multifocal versus corpus callosum) showed a trend. Grade 3 4 toxicities included thrombocytopenia (n = 11) and neutropenia (n = 7) for both pre- and post-rt chemotherapy. Four patients required platelet transfusions. No patient discontinued treatment because of toxicity. Conclusions: The combination of BCNU plus temozolomide as neo-adjuvant therapy in inoperable GBM exhibited promising activity with a good safety profile and warrants further evaluation. Key words: 3-bis(2-chloroethyl)-1-nitrosourea (BCNU), chemotherapy neo-adjuvant strategy, glioblastoma, inoperable brain tumor, temozolomide Introduction Glioblastoma multiforme (GBM) represents the most frequent and aggressive type of primary malignant brain tumor, which occurs in 4 6 per 100 000 people. At present, median survival is only 9 12 months. The standard treatment for GBM is surgery, when feasible, followed by radiotherapy (RT) with or without chemotherapy [1]. Patients for whom debulking surgery is not feasible have a poor prognosis. *Correspondence to: Dr O. Chinot, Unité de Neuro-Oncologie, Service de Neurochirurgie, Hôpital de la Timone, 264 rue Saint Pierre, 13385 Marseille cedex 05, France. Tel: +33-4-91-38-65-69; Fax: +33-4-91-38-73-48; E-mail: olivier.chinot@mail.ap-hm.fr The modest benefit of chemotherapy in patients with GBM has been widely documented and likely reflects the relative insensitivity of GBM to cytotoxic agents. The value of adjuvant chemotherapy is still under discussion. Numerous cooperative group trials have failed to show a statistically significant survival benefit when adjuvant nitrosourea-based chemotherapy was added to conventional RT [2 5]. However, a meta-analyses of 12 randomized trials concluded that chemotherapy adds to the benefit of RT in patients with highgrade gliomas [6]. In addition, a recent randomized, phase III trial conducted by the European Organization for Research and Treatment of Cancer (EORTC) and the National Cancer Institute of Canada (NCIC) demonstrated a statistically significant survival benefit compared with RT alone for patients q 2005 European Society for Medical Oncology

1178 with GBM who were treated with temozolomide concurrently with RT followed by adjuvant temozolomide [7]. These studies suggest that chemotherapy has a role in the treatment of newly diagnosed GBM. Chemotherapy is most commonly administered after RT. However, alternative options include concomitant administration of chemotherapy and RT as reported in the EORTC/NCIC trial [7], or administration of chemotherapy before RT (i.e. neo-adjuvant chemotherapy). Studies investigating neo-adjuvant treatment provide an opportunity to test and select new and effective chemotherapy regimens because the true objective response to chemotherapy alone can be observed. Many studies have tested this approach using either single-agent or combination regimens administered for one to four cycles before RT, and have demonstrated objective response rates ranging from 23% to 54% [8 14]. Both carmustine (BCNU) and temozolomide have demonstrated activity in malignant gliomas and are widely used for the treatment of GBM. Concomitant treatment with temozolomide and RT demonstrated a median survival of 16 months in a phase II trial [15] and a median survival of 15 months in a phase III trial [7]. Temozolomide has also been investigated as neo-adjuvant therapy, before RT, and in this setting produced response rates of 42% to 51% [10, 14]. In addition to these efficacy data, the safety profile of temozolomide is attractive (<5% grade 4 adverse events), which opens the possibility of combination therapy. The rationale for combining BCNU and temozolomide is based on the observed synergic activity of these drugs in preclinical models [16]. This may be attributed, at least in part, to the capacity of temozolomide to deplete intracellular levels of O 6 -alkylguanine-dna alkyltransferase (AGAT) [17], the DNA repair enzyme involved in resistance to BCNU. The strong relationship between AGAT expression and BCNU resistance suggests that modulation of nitrosourea resistance by depletion of AGAT is a reasonable therapeutic strategy [18, 19]. Unfortunately, preclinical models have not been able to clearly discern the optimal sequence and schedule of administration of BCNU and temozolomide to minimize toxicity and maximize antitumor activity [16, 20]. Various schedules have been evaluated clinically and included administration of both drugs on day 1 as proposed by Plowman et al. [16], but this regimen resulted in significant toxicity without evidence of any additive antitumor activity in phase I and phase II studies [21 23]. An alternative schedule involves administration of temozolomide for five consecutive days with BCNU administered on either day 1 or day 5. This was evaluated in a randomized, phase I study, and the results demonstrated sequence-dependent toxicity. Administration of BCNU on day 1 was better tolerated and resulted in a higher tolerated dose of both drugs, as well as a three-fold decrease in AGAT activity [24]. This schedule also demonstrated significant antitumor activity in gliomas, ovarian cancer and sarcomas. These observations are consistent with preclinical models suggesting that administration of BCNU before temozolomide was less toxic and more effective; therefore, we used this schedule of administration in the current study. The objectives of this study were to evaluate the antitumor activity and tolerability of neo-adjuvant BCNU + temozolomide in patients with unresectable, newly diagnosed GBM. This is a patient population for which optimal treatment has not been defined and prognosis remains extremely poor. Patients and methods Newly diagnosed patients with unresectable, histopathologically proven GBM were evaluated for participation in this prospective monocentric study. Histology was obtained by stereotactic biopsy and reviewed by a second pathologist for all patients, and graded according to World Health Organization classification. Eligible patients had to satisfy the following criteria: (i) newly diagnosed GBM not amenable to surgical excision; (ii) no prior radiotherapy or chemotherapy; (iii) 18 70 years of age; (iv) Karnofsky performance status (KPS) of >_ 60; (v) measurable disease based on contrast enhancing lesion on magnetic resonance imaging (MRI); and (vi) adequate hematological, renal and hepatic function. Patients who were pregnant or breast-feeding were ineligible. The protocol was approved by the local medical ethics board, and all patients provided written informed consent. Treatment plan Chemotherapy was administered for up to four cycles, repeated every 42 days, followed by RT. Each cycle included temozolomide (110 mg/m 2 )on day 1 followed 2 h later by BCNU (150 mg/m 2 ) infused over 2 h, and temozolomide (110 mg/m 2 ) on days 2 5. Prophylactic antiemetics using 5-hydroxytryptamine-3 were systematically administered from day 1 to day 5. RT was delivered after four cycles in case of response or stable disease, or earlier in case of tumor progression after the neo-adjuvant chemotherapy period. Focal radiotherapy using a linear accelerator was delivered once daily at 2 Gy per fraction, 5 days/week for a total of 60 Gy. Treatment volumes were determined on the basis of initial gadoliniumenhanced MRI (Gd-MRI) of the brain, and included contrast enhancing lesions plus a 2- to 3-cm margin, depending on tumor volume and location. One month after completion of RT, chemotherapy using the same schedule of BCNU and temozolomide combination was administered for up to four additional cycles, or until progression, unacceptable toxicity or patient refusal. Anticonvulsants were administered as needed, and corticosteroids were used at and adjusted to the optimal dose according to the patient s neurological condition. Patient evaluation Baseline evaluations were performed within 14 days (28 days for imaging) from study entry and included complete medical history, physical and neurological examination, determination of KPS and Mini-Mental Status (MMS), hematology and chemistry assessments, and Gd-MRI. During chemotherapy, complete blood counts were checked weekly and blood chemistry was checked every 3 weeks. A physical and neurological examination, KPS, MMS if appropriate, and a Gd-MRI of the brain were performed after each 42-day cycle for the first four cycles, and then before RT and 1 month after completion of RT. During maintenance chemotherapy, clinical evaluation was performed after each cycle, with Gd-MRI performed at least every two cycles.

1179 Statistical analysis The primary end point of the study was response to treatment. Response was evaluated every cycle for the first four cycles, 4 weeks after completion of radiotherapy, and thereafter at least every two cycles. Response was evaluated according to the Macdonald criteria [25] and based on consecutive Gd-MRI under stable steroid use for 7 days before scan. For evaluation, the slice with the largest tumor area was used. Tumor size was defined as the product of the two largest perpendicular tumor diameters. Complete response (CR) was defined as disappearance of all contrastenhancing tumor on two subsequent scans at least 1 month apart, and the patient being off steroids and neurologically stable or improved. Partial response (PR) was defined as a >_ 50% reduction in cross-sectional contrast-enhancing tumor area on two subsequent scans at least 1 month apart, steroids stable or decreased and neurologically stable or improved. Progressive disease (PD) was defined as a >25% increase in crosssectional contrast-enhancing tumor area, new tumor on MRI or neurological deterioration, and steroids stable or increased. All other situations maintained for at least 3 months were considered stable disease (SD). All patient scans in which a response (CR or PR) was reported were centrally reviewed by an independent expert (K.H.X). Secondary end points were overall survival (OS), progression-free survival (PFS) and safety of treatment. OS was measured from the date of study entry to the date of death or last follow-up. PFS was measured from the date of study entry to the first sign of radiological or clinical progression, whichever came first. Survival and PFS are reported separately for the intention-to-treat (ITT) and eligible population. Estimates of OS and PFS were derived by the Kaplan Meier method. Log-rank tests were used to compare OS and PFS between groups. Hematological and nonhematological toxicities were assessed using the National Cancer Institute Common Toxicity Criteria. Results Patients Forty patients were enrolled between March 2000 and May 2002, and these patients constitute the ITT population. Three patients were ineligible (two because of inappropriate histological material and one because of non-measurable disease), 39 patients were assessable for response, and all patients were assessable for survival and safety. Patient demographics and baseline disease characteristics are outlined in Table 1. The median age was 61 years and one-third of the patients had a KPS of 60. Surgery was limited to biopsy for all patients. The median time from diagnosis to chemotherapy was 13 days (range 5 32). Treatment delivery Treatment delivered is reported in Table 2. There were no discontinuations because of toxicity. Sixty per cent of the patients completed the planned four cycles of BCNU + temozolomide prior to RT, and the remaining patients experienced early disease progression after one (16%), two (16%) or three (8%) cycles. Dose reduction was requested in 31 of 162 (19%) cycles administered. Cycle 1 was administered at full dose for all patients; therefore, dose reduction took place in 31 of the remaining 122 cycles delivered. Doses were reduced for 13 of 34 patients who received more than one cycle. Table 1. Patient demographics and clinical characteristics Characteristic Total number of patients 40 Age, years Median 61 Range 36 70 Sex, n (%) Male 27 (68) Female 13 (32) Karnofsky performance status, n (%) 60 15 (37.5) 70 80 25 (67.5) Location, n (%) Lobar 23 (58) Multifocal 10 (25) Corpus callosum 7 (17) Tumor size, cm 2 Median ± SD 12.6 ± 9.4 Mean 14.5 Range 2.2 37.4 Stereotactic biopsy, n (%) 40 (100) Time from diagnosis to treatment, days Median 13 Range 5 32 Histology review: GBM, n (%) 38 (95) SD, standard deviation; GBM, glioblastoma multiforme. Table 2. Treatment delivered for 37 eligible patients Treatment n (%) Neo-adjuvant chemotherapy 37 Completed four cycles before RT 22 (60) Discontinued due to PD 15 (40) After one cycle 6 After two cycles 6 After three cycles 3 Radiotherapy RT not started or not completed 9 (24) Multifocal 4 (9) Corpus callosum 3 (7) Lobar 2 (21) Maintenance chemotherapy 14 Discontinuation after five cycles 5 (14) six cycles 3 (8) seven cycles 3 (8) Completed eight cycles 3 (8) RT, radiotherapy; PD, progressive disease.

1180 Therapeutic RT was completed in 31 (78%) patients; only two patients were unable to complete RT because of early major clinical and radiological disease progression. In the remaining seven cases, RT was not given because of the patient s clinical status and anatomical considerations (e.g. large volumes to be irradiated, multifocal lesions in four cases and large corpus callosum tumors in three cases). Fourteen patients received maintenance chemotherapy after RT, including five patients who completed five cycles, three who completed six cycles, three who completed seven cycles, and three who completed eight cycles of BCNU + temozolomide. Tumor response By ITT analysis, 17 [42.5%; 95% confidence interval (CI) 27% to 58%] patients achieved an objective response to neoadjuvant BCNU + temozolomide, including two (5%) patients with a CR and 15 (37.5%) patients with a PR. In addition, nine (22.5%) patients had SD and 14 (35%) had PD. In the eligible population (n = 37), the objective response rate was 46% (95% CI 30% to 62%), including 5% CR and 41% PR. All these responses were confirmed at least 6 weeks apart and were maintained at least until the start of RT. Most of the responses (16 of 17) occurred during the first three cycles (i.e. within the first 4.5 months); objective responses were observed after one cycle in five patients, after two cycles in four patients and after three cycles in seven patients. Objective response was observed in two of seven (29%) patients with corpus callosum tumors, in four of 10 (40%) patients with multifocal tumors and in 11 of 23 (48%) patients with lobar tumors. Survival Median follow-up was 18.8 months (range 6 33.5) for the entire population. Based on ITT analysis, median OS was 12.7 months (Figure 1), and the 12- and 18-month survival rates were 54% (95% CI 37.6% to 70.3%) and 20% (95% CI 5.7% to 34.9%), respectively. Results were similar in the ITT and eligible patient populations (Table 3). At the time of analysis, a total of 10 (25%) patients were alive. The median follow-up of these patients was 12.7 months (range 6 29). Based on ITT analysis, median PFS was 7.4 months (Figure 2) [26]. Among patients with an objective response (CR or PR), median OS was 16.6 months and median PFS was 10.3 months (Table 3). Prognostic factors Of the potential factors affecting prognosis (i.e. age, KPS, tumor size and tumor location), only age appeared to significantly influence survival, with an OS of 16.6 months for patients <_ 50 years of age and 9.8 months for patients >50 years of age (P = 0.032). However, tumor location (i.e. lobar versus multifocal versus corpus callosum) also showed a trend toward influencing survival, but did not reach statistical significance. An analysis of patients according to the recursive partitioning analysis (RPA) prognostic classes, which are based on the Radiation Therapy Oncology Group database of patients with malignant gliomas [27], revealed that patients in this study who were classified as RPA V (n = 18) had longer survival rates (median OS 14 months; 2-year survival rate 15.1%; 95% CI 0% to 39.3%) than patients classified as RPA VI (n = 11) (median OS 5.3 months; 2-year survival rate 0%). Safety A total of 162 cycles of BCNU + temozolomide were administered. Treatment was well tolerated and was administered at full dose to 27 of 40 (68%) patients. No patient died of toxicity. There were 32 grade 3 or 4 adverse events, including 13 grade 4 adverse events all related to myelosuppression (Table 4). These grade 4 events affected only six patients in the study. Platelet transfusions were administered seven times in four patients. Hospitalization for toxicity was required for three patients. No clinically relevant pulmonary complication Figure 1. Kaplan Meier estimate of overall survival in months from time of study entry.

Table 3. Neo-adjuvant chemotherapy with BCNU + temozolomide n (%) Median PFS (months) Median OS (months) ITT 40 7.4 12.7 Eligible 37 7.9 12.0 ITT (CR+PR) 17 (42.5) 10.3 16.6 SD 9 (22.5) 8.2 14.5 PD 14 (35.0) 2.6 4.4 BCNU, carmustine; PFS, progression-free survival; OS, overall survival; ITT, intention-to-treat population; CR, complete response; PR, partial response; SD, stable disease; PD, progressive disease. was observed, and carbon monoxide diffusion capacity performed on 12 patients did not identify any significant alteration. Discussion The primary goal of this study was to evaluate the efficacy of the BCNU + temozolomide combination as neo-adjuvant therapy in GBM prior to RT. In the neo-adjuvant setting, the goal of chemotherapy is to reduce tumor burden, which may improve response to subsequent RT. Therefore, objective response is the most relevant end point in this setting. This regimen demonstrated promising antitumor activity, with an objective response rate of 42.5% (ITT analysis), which is particularly encouraging considering that this was a poor prognosis group of patients with unresectable GBM and considering that all responses were confirmed 1 month apart and were independently reviewed. The observed response rate in this study compares favorably with that of other agents and regimens that have been explored as neo-adjuvant therapy in patients with high-grade gliomas (Table 5). In this setting, reported response rates to chemotherapy in patients with mainly GBM varied from 16% to 54% [8 14, 28, 29]. The combination of BCNU plus Table 4. Major grade 3 or 4 toxicity with combination of BCNU + temozolomide Number of events per cycle (n = 162) Grade 3 Grade 4 Constipation 2 0 Nausea 3 0 Emesis 2 0 Fatigue 5 0 Thrombocytopenia 5 9 Neutropenia 1 4 Anemia 1 0 1181 Toxicity was graded according to the National Cancer Institute Common Toxicity Criteria. BCNU, carmustine. cisplatin has been the most frequently evaluated regimen. Surprisingly, four studies that explored a similar monthly schedule (BCNU 40 mg/m 2 and cisplatin 40 mg/m 2, both on days 1 3) in unselected patient populations demonstrated response rates that varied widely, from 16% to 54% [8, 13, 28, 29]. Two other studies tested a platinum-based combination that included etoposide (VP16), with or without fotemustine, which demonstrated response rates of 27% and 24%, respectively [11, 12]. Finally, two studies testing single-agent temozolomide using the standard 5-day schedule have reported high response rates of 42% and 51% in patients with GBM [10, 14]. Although the response rate achieved in the current study with the combination of BCNU + temozolomide is similar to that achieved with temozolomide alone, the majority (60% to 70%) of patients in the two studies of single-agent temozolomide had subtotal resection of their tumors prior to chemotherapy, potentially affecting the response rate. Figure 2. Kaplan Meier estimate of progression-free survival in months from time of study. Data from Chinot et al. [26].

1182 Table 5. Clinical studies of chemotherapy administered before radiotherapy in GBM Study Schedule n GBM (%) STRB (%) OR (%) PFS (months) OS (months) Recht 1990 [8] BCNU + CDDP 30 80 10 16 12 14 Kirby 1996 [9] PCV 22 100 NR 5.6 10 Grossman 1997 [28] BCNU + CDDP 52 88 29 42 NR 12.9 Friedman 1998 [10] Temozolomide 33 100 51 NR NR Jeremic 1999 [11] CB + VP16 45 77 29 24 12 14 Frenay 2000 [12] FOTE + CDDP + VP16 33 100 100 27 NR 10 Gilbert 2000 [13] BCNU + CDDP 47 100 22 NR NR Dazzi 2000 [29] BCNU + CDDP 18 83 54 NR 9 Gilbert 2002 [14] Temozolomide 36 100 39 42 3.9 13.2 This study BCNU + temozolomide 40 100 100 42.5 7.4 12.7 GBM, glioblastoma multiforme; STRB, stereotactic biopsy only; OR, objective response rate; PFS, median progression-free survival; OS, median overall survival; CDDP, cisplatinum; BCNU, carmustine; PCV, procarbazine, CCNU and vincristine; CB, carboplatin; VP16, etoposide; FOTE, fotemustine; NR, not reported. In assessing the effect of neo-adjuvant BCNU + temozolomide on survival, it must be noted that patients in the current study presented with several factors associated with poor prognosis for survival (i.e. unresectable and multifocal disease, median age of 61 years and median KPS of 70). In fact, most patients (73%) in our study would be classified in RPA class V or VI. In the original cohort published by Curran et al. [27], patients in class V (n = 395) had a median survival of 8.9 months, and patients in class VI (n = 263) had a median survival of 4.6 months. In comparison, patients in the current study who would be classified as RPA class V (n = 18) had a median survival of 14 months, and patients in RPA class VI (n = 11) had a median survival of 5.3 months. Therefore, although direct comparisons between these patient populations are difficult because of differences in selection criteria and changes in histological criteria, our data suggest that neo-adjuvant BCNU + temozolomide may provide a survival benefit in these poor prognosis patients. In addition, the overall median survival of 12.7 months reported in the current study compares favorably with previous studies in patients with inoperable GBM that reported median survivals of only 7 months with conventional RT [30], 8.6 months with a combination of RT plus carboplatin and nicotinamide [31] and 10 months with neo-adjuvant chemotherapy (fotemustine, cisplatin and VP16) followed by conventional RT, as reported by Frenay et al. [12] (Table 5). Our results also compare favorably with the combination of temozolomide and RT reported by Stupp et al. [7, 15]. In these trials, only 20% of patients presented with inoperable tumors, and median survival in this subset was only 5.3 months in the phase II trial [15] and 9.4 months in the phase III trial [7]. PFS and OS in the current study also appear better than those seen in studies of single-agent temozolomide. For example, in the recent study reported by Gilbert et al. [14] (Table 5), median PFS was 3.9 months and median OS was 13.2 months. These values are not significantly improved compared with historical controls considering that only 39% of patients had inoperable tumors and 86% of patients had KPS >_80. In comparison, we have shown that BCNU + temozolomide in patients with unresectable tumors and poor performance status yielded a median PFS of 7.4 months in the ITT population and a median survival of 12.7 months. Moreover, among patients achieving an objective response, median PFS was 10.3 months. Ultimately, these considerable delays in disease progression may translate into a survival benefit. In the poor prognosis patient population enrolled in this study, only age demonstrated prognostic significance, whereas tumor location tended to influence survival but not significantly. These results confirm that GBM constitutes a heterogenous group of tumors and suggest that a significant proportion of tumors are chemosensitive independent of age, KPS or surgery. These results are in accordance with the meta-analysis conducted by Stewart [6], in which the benefit of adjuvant chemotherapy was independent of these prognostic factors. This underlines the necessity to determine appropriate molecular markers of sensitivity or resistance to treatment in order to better characterize tumors and determine the most effective treatment strategy for each individual patient. Taken together, the results of this study suggest that the combination of BCNU and temozolomide, using the particular schedule investigated, may offer some benefit over singleagent temozolomide or BCNU. The treatment benefit associated with this combination may be related to modulation of AGAT activity consistent with preclinical and phase I clinical data, as suggested by Hammond et al. [24]. However, there remains some uncertainty about the clinical relevance of measuring AGAT activity in peripheral blood mononuclear cells, the correlation between AGAT depletion and sensitivity to methylating agents, and the relative impact of AGAT levels compared with other resistance mechanisms such as aberrant mismatch repair. Other strategies to maximize AGAT depletion with a continuous daily dosing schedule of temozolomide are under evaluation and will include correlation to AGAT levels in tumor tissue [24, 32]. In the current study, because of the limited tumor material obtained from stereotactic biopsies, these correlative studies could not be performed.

1183 Table 6. Toxicity of alternative BCNU temozolomide schedules Study Schedule Toxicity per cycle (%) Dose reduction beyond cycle 1 (%) Grade 3 Grade 4 Per cycle Per patient This study BCNU 150 mg/m 2 at day 1 + TMZ 110 mg/m 2 from 12 8 25 32 0 day 1 to day 5 Chang 2004 [23] BCNU 150 mg/m 2 + TMZ 550 mg/m 2 on day 1 46 44 NR 58 22 Prados 2004 [22] BCNU 150 mg/m 2 + TMZ 550 mg/m 2 on day 1 36 32 67 NR NR BCNU, carmustine; TMZ, temozolomide; NR, not recorded. Treatment interruption before cycle 4 for toxicity (%) Our results also showed that administration of BCNU on day 1 of each 5-day course of temozolomide was well tolerated; 60% of patients completed the planned four cycles of neo-adjuvant therapy, and none discontinued because of toxicity. Moreover, dose reductions were required in only 25% of cycles beyond cycle 1. In contrast, when neo-adjuvant BCNU and temozolomide were both administered on day 1, substantial toxicity was observed (Table 6) [22, 23]. In one study, 32% of patients developed grade 4 myelosuppression and dose reduction was required in 67% of cycles beyond cycle 1 [22]. In a similar study, 46% of patients experiencing grade 3 or 4 thrombocytopenia, and there were two toxic deaths [23]. Of the 41 eligible patients, nine (22%) patients discontinued early because of toxicity and 14 of 24 patients who completed the four planned cycles required dose reductions. In conclusion, the combination of BCNU + temozolomide before RT, using a schedule that improved compliance compared with the 1-day schedule, exhibited promising clinical activity in unresectable GBM. Therefore, this regimen warrants further evaluation in malignant gliomas. While the combination of RT and temozolomide, as described by Stupp et al. [7], may become the new standard for adjuvant therapy following surgical resection, some subgroups of patients, such as those with unresectable tumors, may benefit from neo-adjuvant therapy to reduce tumor burden before standard adjuvant RT. This provides an opportunity to evaluate new treatment strategies. Acknowledgements Claire Senay is acknowledged for shaping the manuscript. This work was supported by the Foundation Lionel Perrier and GRENONE. References 1. Chinot OL. Should radiotherapy be standard therapy for brain tumors in the elderly? Cons Semin Oncol 2003; 30 (Suppl): 68 71. 2. Green SB, Byar DP, Walker MD et al. Comparisons of carmustine, procarbazine, and high-dose methylprednisolone as additions to surgery and radiotherapy for the treatment of malignant glioma. Cancer Treat Rep 1983; 67: 121 132. 3. Chang CH, Horton J, Schoenfeld D et al. Comparison of postoperative radiotherapy and combined postoperative radiotherapy and chemotherapy in the multidisciplinary management of malignant gliomas. A joint Radiation Therapy Oncology Group and Eastern Cooperative Oncology Group study. Cancer 1983; 52: 997 1007. 4. Shapiro WR, Green SB, Burger PC et al. Randomized trial of three chemotherapy regimens and two radiotherapy regimens in postoperative treatment of malignant glioma. Brain Tumor Cooperative Group trial 8001. J Neurosurg 1989; 71: 1 9. 5. Medical Research Council Brain Tumor Working Party. Randomized trial of procarbazine, lomustine, and vincristine in the adjuvant treatment of high-grade astrocytoma: a Medical Research Council trial. J Clin Oncol 2001; 19: 509 518. 6. Stewart LA. Chemotherapy in adult high-grade glioma: a systematic review and meta-analysis of individual patient data from 12 randomised trials. Lancet 2002; 359: 1011 1018. 7. Stupp R, Mason WP, Van Den Bent MJ et al. Concomitant and adjuvant temozolomide (TMZ) and radiotherapy (RT) for newly diagnosed glioblastoma multiforme (GBM). Conclusive results of a randomized phase III trial by the EORTC Brain & RT Groups and NCIC Clinical Trials Group. Proc Am Soc Clin Oncol 2004; 23: 1 (Abstr 2). 8. Recht L, Fram RJ, Strauss G et al. Preirradiation chemotherapy of supratentorial malignant primary brain tumors with intracarotid cisplatinum (CDDP) and i.v. BCNU. A phase II trial. Am J Clin Oncol 1990; 13: 125 131. 9. Kirby S, Macdonald D, Fisher B et al. Pre-radiation chemotherapy for malignant glioma in adults. Can J Neurol Sci 1996; 23: 123 127. 10. Friedman HS, McLendon RE, Kerby T et al. DNA mismatch repair and O 6 -alkylguanine-dna alkyltransferase analysis and response to Temodal in newly diagnosed malignant glioma. J Clin Oncol 1998; 16: 3851 3857. 11. Jeremic B, Shibamoto Y, Grujicic D et al. Pre-irradiation carboplatin and etoposide and accelerated hyperfractionated radiation therapy in patients with high-grade astrocytomas: a phase II study. Radiother Oncol 1999; 51: 27 33. 12. Frenay M, Lebrun C, Lonjon M et al. Up-front chemotherapy with fotemustine (F)/cisplatin (CDDP)/etoposide (VP16) regimen in the treatment of 33 non-removable glioblastomas. Eur J Cancer 2000; 36: 1026 1031. 13. Gilbert M, O Neill A, Grossman S et al. A phase II study of preradiation chemotherapy followed by external beam radiotherapy for the treatment of patients with newly diagnosed glioblastoma multiforme: an Eastern Cooperative Oncology Group study (E2393). Neuro-oncol 2000; 47: 145 152. 14. Gilbert MR, Friedman HS, Kuttesch JF et al. A phase II study of temozolomide in patients with newly diagnosed supratentorial malignant glioma before radiation therapy. Neuro-oncol 2002; 4: 261 267. 15. Stupp R, Dietrich P-Y, Kraljevic SO et al. Promising survival for patients with newly diagnosed glioblastoma multiforme treated with

1184 concomitant radiation plus temozolomide followed by adjuvant temozolomide. J Clin Oncol 2002; 20: 1375 1382. 16. Plowman J, Waud WR, Koutsoukos AD et al. Preclinical antitumor activity of temozolomide in mice: efficacy against human brain tumor xenografts and synergism with 1,3-bis(2-chloroethyl)-1-nitrosourea. Cancer Res 1994; 54: 3793 3799. 17. Lee SM, Thatcher N, Crowther D, Margison GP. Inactivation of O6-alkylguanine-DNA alkyltransferase in human peripheral blood mononuclear cells by temozolomide. Br J Cancer 1994; 69: 452 456. 18. Pegg AE. Mammalian O6-alkylguanine-DNA alkyltransferase: regulation and importance in response to alkylating carcinogenic and therapeutic agents. Cancer Res 1990; 50: 6119 6129. 19. Gerson SL. Regeneration of O6-alkylguanine-DNA alkyltransferase in human lymphocytes after nitrosourea exposure. Cancer Res 1988; 48: 5368 5373. 20. Mitchell RB, Dolan ME. Effect of temozolomide and dacarbazine on O6-alkylguanine-DNA alkyltransferase activity and sensitivity of human tumor cells and xenografts to 1,3-bis(2-chloroethyl)-1-nitrosourea. Cancer Chemother Pharmacol 1993; 32: 59 63. 21. Schold SC Jr, Kuhn JG, Chang SM et al. A phase I trial of 1,3-bis(2- chloroethyl)-1-nitrosourea plus temozolomide: a North American Brain Tumor Consortium study [published erratum appears in Neurooncol 2001; 3 123]. J Neuro-ncol 2000; 2: 34 39. 22. Prados MD, Yung WK, Fine HA et al. Phase 2 study of BCNU and temozolomide for recurrent glioblastoma multiforme: North American Brain Tumor Consortium study. Neuro-oncol 2004; 6: 33 37. 23. Chang SM, Prados MD, Yung WKA et al. Phase II study of neoadjuvant 1, 3-bis (2-chloroethyl)-1-nitrosourea and temozolomide for newly diagnosed anaplastic glioma: a North American Brain Tumor Consortium Trial. Cancer 2004; 100: 1712 1716. 24. Hammond LA, Eckardt JR, Kuhn JG et al. A randomized phase I and pharmacological trial of sequences of 1,3-bis(2-chloroethyl)-1-nitrosourea and temozolomide in patients with advanced solid neoplasms. Clin Cancer Res 2004; 10: 1645 1656. 25. Macdonald DR, Cascino TL, Schold SC Jr, Cairncross JG. Response criteria for phase II studies of supratentorial malignant glioma. J Clin Oncol 1990; 8: 1277 1280. 26. Chinot O, Barrie M, Couprie C et al. Preradiation combination of temozolomide (TMZ) and BCNU as primary treatment before radiotherapy (RT) in inoperable newly diagnosed glioblastoma multiforme (GBM) [poster]. Presented at the 39th Annual Meeting of the American Society of Clinical Oncology 31 May to 3 June 2003, Chicago, IL, USA. Proc Am Soc Clin Oncol: 108 (Abstr 431). 27. Curran WJ Jr, Scott CB, Horton J et al. Recursive partitioning analysis of prognostic factors in three Radiation Therapy Oncology Group malignant glioma trials. J Natl Cancer Inst 1993; 85: 704 710. 28. Grossman SA, Wharam M, Sheidler V et al. Phase II study of continuous infusion carmustine and cisplatin followed by cranial irradiation in adults with newly diagnosed high-grade astrocytoma. J Clin Oncol 1997; 15: 2596 2603. 29. Dazzi C, Cariello A, Giannini M et al. A sequential chemo-radiotherapeutic treatment for patients with malignant gliomas: a phase II pilot study. Anticancer Res 2000; 20: 515 518. 30. Coffey RJ, Lunsford LD, Taylor FH. Survival after stereotactic biopsy of malignant gliomas. Neurosurgery 1988; 22: 465 473. 31. Simon JM, Noël G, Chiras J et al. Radiotherapy and chemottherapy with or without carbogen and nicotinamide in inoperable biopsy-proven glioblastoma multiforme. Radiother Oncol 2003; 67: 45 51. 32. Friedman HS, Dolan ME, Pegg AE et al. Activity of temozolomide in the treatment of central nervous system tumor xenografts. Cancer Res 1995; 55: 2853 2857.